Skip to main content

ANSVIS v1.2.0.26

New Features​

Face Recognition​

Performance And Task Execution​

  • Added the Active Engine setting to control the number of AI tasks running in parallel for resource optimization.
  • Added Input Trigger to define task execution conditions. This feature enables automated event capture and processing based on external signals. It is available for pre-built and custom tasks.

Rules And Triggers​

  • Added the IF-THEN rule join condition.
  • Added Line ROI Gap to increase line thickness and reduce false detections when an object moves near the line. It is available for pre-built and custom tasks.
  • Added a Fully Serialized custom HTTP trigger format that separates each detected object in an event into an individual HTTP trigger.

Pre-Built AI Tasks And Models​

  • Added model sensitivity and a dedicated smoke score for the Fire & Smoke Detection task.
  • Added the Improper Object Placement task to detect objects placed in restricted areas or locations where they may cause obstructions.
  • Added the ANS Objects model, which includes objects for monitoring.

Improvements​

Face Recognition​

  • Updated face recognition behavior when a task is configured to trigger only one event per ID. If an unknown face is later recognized, ANSVIS now generates a second event so the recognized face is captured.
  • Added a required user ID for new users. If the user ID field is empty, ANSVIS automatically generates an ID that can be modified later.
  • Updated the face labeling process in Face Recording Tasks.
  • Improved face detection accuracy and processing speed.

Performance​

  • Improved overall system responsiveness and resource utilization, including lower CPU and GPU usage, a reduced memory footprint, and improved throughput for high-density camera deployments.
  • Improved ALPR detection speed and accuracy.

Tasks And Documentation​

  • Updated the description for each pre-built AI task. Select the i icon next to the task name to view its description.
  • Updated the pre-built AI task configuration layout.
  • Updated image snapshot behavior. ANSVIS now takes a few additional seconds to help capture a good image.
  • Made the AIBOX mobile app available worldwide on the App Store and Google Play.

Triggers​

  • Updated the HTTP Send Test function to run from ANSVIS Server instead of ANSVIS.

Bug Fixes​

Performance And Stability​

  • Fixed an issue where cameras continued streaming in the background after a task was stopped.

Background Processes​

  • Fixed an issue where the Path rule did not detect objects for some movement patterns.
  • Fixed incorrect custom HTTP trigger formatting when an object did not have an attribute.
  • Fixed an issue where email trigger attachments did not include bounding boxes.